Title：Optimizing information quality in data collection from mobile sensor networks to mobile crowdsensing
Speaker：Associate Prof. Edith Ngai
Host：Prof. Wendong Wang
Time：December 22nd,2014 10:00-12:00 AM
Location：New Research Building Room 510
Abstract：
With advancements of mobile devices, mobile users can collect sensing data from their surrounding wireless sensors using Bluetooth or NFC. Due to the limited battery and contact time between the sensors and the mobile devices, it is crucial to prioritize sensing data to maximize information utility in opportunistic data collection. We present a novel context-aware sensing data dissemination framework that maximizes information utility by considering sensing context, including sensing type, locality, time-to-live, mobility and user interests. We evaluated our solution by both analysis and simulations, and showed that it can provide high information utility for mobile users with low communication overhead.
Apart from relaying sensing data, the sensing capability of mobile devices enables them to act as mobile sensors to support many crowdsensing applications. Incentive allocation is an important research topic in mobile crowdsensing. Unfortunately, existing incentive approaches usually include inefficient negotiation procedures. We propose a novel incentive allocation mechanism to address the problem. Our solution is able to encourage participation and allocate incentives dynamically to achieve accurate sensing results.
Finally, we would like to present a mobile social crowd sensing application for safe driving. In this work, we investigate Mood-Fatigue Analyzer (MFA), which employs multidimensional methods to obtain drivers' real-time mood and fatigue information by sensors inside and outside the vehicles. Besides promoting safe driving with integrated sensors, the MFA also enables communication between drivers in a social context via the cloud.
SKL-NST
December 16th,2014
